Recent studies have found rapid and highly efficacious antidepressant effects of a single ketamine infusion in treatment-resistant depression (TRD). However, a single infusion appears insufficient to maintain response as most patients return to previous depressive state within a week. The strategy of multiple infusions to increase efficacy and sustain antidepressant effects has not yet been systematically evaluated in an RCT. The proposed study is a one-center, interventional, efficacy study designed to determine antidepressant outcomes of serial ketamine infusions compared to a single ketamine infusion among veterans with TRD. We have hypothesized that six infusions will be superior to a single infusion of ketamine in both decreasing severity of depressive symptoms and maintaining response. Participants will be male/female veterans (18 to 75 years old) of any era or military background who suffer from TRD defined as failure to achieve remission from at least 2 antidepressant trials of different pharmacological classes. Potential participants will be recruited from Mental Health clinics and screened for eligibility using a two stage process (phone/chart review, followed by interview). Exclusion criteria includes post-traumatic stress disorder, psychosis-related disorder, bipolar disorder, alcohol/substance use disorder 6 months prior to screening; unstable medical illness; serious/imminent suicidal/homicidal risk; Parkinson's disease, dementia, seizures; traumatic brain injury; contraindicated medications (e.g., MAO inhibitors, barbiturates); received ECT during current episode; pregnancy/nursing. Baseline assessments will be completed 1-2 weeks prior to starting treatment. Participants will be randomly assigned to one of two parallel treatment conditions: 1) six ketamine infusions at 0.5 mg/kg or 2) single ketamine infusion at 0.5 mg/kg preceded by five midazolam infusions at 0.045 mg/kg. Midazolam was chosen as an active placebo given similar pharmacokinetics and dissociative effect profile to ketamine. Each intervention will be provided for a total of 12-day infusion-phase on a Monday-Wednesday-Friday schedule. The, follow-up visits will occur at weekly intervals for the first 4 weeks, at 2-week intervals for the next 8 weeks, and at 4-week intervals for the remaining 12 weeks or until relapse. The primary end point is the Montgomery- ?sberg Depression Rating Scale (MADRS) score 24 hours following the last infusion where the peak antidepressant effects of ketamine occur. Secondary outcomes for the treatment phase include remission defined by MADRS<10, and response defined as a reduction in the baseline MADRS score e 50%. For the follow-up period, durability of antidepressant response will be defined by time to relapse to a MADRS score <50% of baseline at that visit. Independent evaluation of depressive symptom severity and potential covariates of antidepressant effect (e.g., pain intensity, level of anxiety) will be ascertained at baseline, at several time points during infusion period, and at follow-up. On the day of infusion, subjects will arrive in the morning after an overnight fast. Hemodynamic measures will be recorded every 10 min for 1 hour beginning 10 min before infusion. Subjects will then receive IV infusion over 40 minutes. Severity of depressive symptoms, pain intensity, level of anxiety will be obtained before and 24 hours after infusion. Acute dissociative effects, manic/hypomanic symptoms, and psychotomimetic effects will be measured 30 minutes before the start of each infusion and at the end of infusion (t0+40 mins) and again at t0+120mins and t0+180mins. The infusion will be discontinued in the event of significant adverse events. Procedures for the subsequent infusions at days 3, 5, 8, 10, and 12 will be identical to those of the first infusion.